CVE-2026-64464 xhci: sideband: fix ring sg table pages leak

In the Linux kernel, the following vulnerability has been resolved: xhci: sideband: fix ring sg table pages leak xhciringtosgtable allocates a temporary pages array and uses it to build the returned sgtable with sgalloctablefrompages. The error paths free the pages array, but the success path...

CVE-2026-64464

Consolidated details show CVE-2026-64464 affects the Linux kernel xhci sideband path: xhci_ring_to_sgtable() allocates a temporary pages array used to build an sg_table via sg_alloc_table_from_pages(). On success, the temporary array isn’t freed, causing a leak each time a sideband client uses an...

CVE-2026-64414 netfilter: handle unreadable frags

In the Linux kernel, the following vulnerability has been resolved: netfilter: handle unreadable frags sashiko reports: When an skb with unreadable fragments such as from devmem TCP, where skbfragsreadableskb returns false is processed by the u32 module, skbcopybits will safely return a negative...

CVE-2026-64414 netfilter: handle unreadable frags

In the Linux kernel, the following vulnerability has been resolved: netfilter: handle unreadable frags sashiko reports: When an skb with unreadable fragments such as from devmem TCP, where skbfragsreadableskb returns false is processed by the u32 module, skbcopybits will safely return a negative...

CVE-2026-64269

In the Linux kernel, the following vulnerability has been resolved: RDMA/rtrs-srv: Bound RDMA-Write length to chunk size in rdmawritesg When the server answers an RTRS READ, rdmawritesg builds the source scatter/gather entry for the IBWRRDMAWRITE that returns data to the peer. Its length is taken...

CVE-2026-64019

A flaw was found in the Linux kernel's NVMe Non-Volatile Memory Express PCI Peripheral Component Interconnect driver. This vulnerability occurs due to a leak in the Direct Memory Access DMA mapping during data setup operations. Specifically, the system fails to properly unmap DMA resources when...

SUSE CVE-2026-64124

In the Linux kernel, the following vulnerability has been resolved: net: devmem: reject dma-buf bind with non-page-aligned size or SG length netdevmembinddmabuf trusts dmabuf-size and sgdmalen to be PAGESIZE multiples without checking: - txvec is sized dmabuf-size / PAGESIZE, and netdevmemgetniov...

CVE-2026-64124

A flaw was found in the Linux kernel's network device memory net: devmem component. This vulnerability allows an attacker to trigger an out-of-bounds read by providing specially crafted Direct Memory Access DMA buffers with non-page-aligned sizes or Scatter-Gather SG list lengths. This could lead...

net: skmsg: preserve sg.copy across SG transforms

...

CVE-2026-64124 net: devmem: reject dma-buf bind with non-page-aligned size or SG length

In the Linux kernel, the following vulnerability has been resolved: net: devmem: reject dma-buf bind with non-page-aligned size or SG length netdevmembinddmabuf trusts dmabuf-size and sgdmalen to be PAGESIZE multiples without checking: - txvec is sized dmabuf-size / PAGESIZE, and netdevmemgetniov...

PYSEC-2026-3172 TensorFlow vulnerable to `CHECK` fail in `CollectiveGather`

Impact When CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack. python import tensorflow as tf arg0=1 arg1=1 arg2=1 arg3=1 arg4=3, 3,3 arg5='auto' arg6=0 arg7='' tf.rawops.CollectiveGatherinput=arg0, groupsize=arg1,...

TensorFlow vulnerable to `CHECK` fail in `CollectiveGather`

ImpactWhen CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack.pythonimport tensorflow as tfarg0=1arg1=1arg2=1arg3=1arg4=3, 3,3arg5='auto'arg6=0arg7=''tf.rawops.CollectiveGatherinput=arg0, groupsize=arg1, groupkey=arg2,...

PYSEC-2026-3320 TensorFlow vulnerable to `CHECK` fail in `CollectiveGather`

Impact When CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack. python import tensorflow as tf arg0=1 arg1=1 arg2=1 arg3=1 arg4=3, 3,3 arg5='auto' arg6=0 arg7='' tf.rawops.CollectiveGatherinput=arg0, groupsize=arg1,...

PT-2026-59789

Impact When CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack. python import tensorflow as tf arg 0=1 arg 1=1 arg 2=1 arg 3=1 arg 4=3, 3,3 arg 5='auto' arg 6=0 arg 7='' tf.raw ops.CollectiveGatherinput=arg 0, group...

PYSEC-2026-990 TensorFlow vulnerable to `CHECK` fail in `CollectiveGather`

Impact When CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack. python import tensorflow as tf arg0=1 arg1=1 arg2=1 arg3=1 arg4=3, 3,3 arg5='auto' arg6=0 arg7='' tf.rawops.CollectiveGatherinput=arg0, groupsize=arg1,...

PT-2026-53523

Out-of-bounds read in gather tree in PaddlePaddle before 2.4. A patch is available in the release/2.4 branch...

SUSE CVE-2026-53133

In the Linux kernel, the following vulnerability has been resolved: RDMA/umem: Fix truncation for block sizes = 4G When the iommu is used the linearization of the mapping can give a single block that is very large split across multiple SG entries. When rdmablockiternext reassembles the split SG...

CVE-2026-53133

A flaw was found in the Linux kernel's RDMA/umem component. When the IOMMU Input/Output Memory Management Unit is used, a very large memory block can be split across multiple scatter-gather SG entries. During the reassembly of these split SG entries, an issue with truncation for block sizes great...

CVE-2026-53133

In the Linux kernel, the following vulnerability has been resolved: RDMA/umem: Fix truncation for block sizes = 4G When the iommu is used the linearization of the mapping can give a single block that is very large split across multiple SG entries. When rdmablockiternext reassembles the split SG...

CVE-2026-53133 RDMA/umem: Fix truncation for block sizes >= 4G

In the Linux kernel, the following vulnerability has been resolved: RDMA/umem: Fix truncation for block sizes = 4G When the iommu is used the linearization of the mapping can give a single block that is very large split across multiple SG entries. When rdmablockiternext reassembles the split SG...